Skip to content

Latest commit

 

History

History
26 lines (17 loc) · 1.28 KB

File metadata and controls

26 lines (17 loc) · 1.28 KB

Clustering-of-Ward-Level-Deprivation-through-the-use-of-Satellite-Imagery

This is the Github repo for our honours project. It is still in progress.

Notebooks

get_image_centers.ipynb: Reading in the coordinates for the Gauteng wards from the geojson shapefile, and then writing the centers for the images to /data/ward_image_centers.csv. These centers will then be used to extract the images.

get_images.ipynb: Using /data/ward_image_centers.csv, this notebook then gets two images from Google Static Maps API for each of the center coordinates - one with the ward overlayed in black and one without. These images are stored in /images/ward_k/ij.png where k is the ward number and ij are the row and column positions, respectively, of the image in the grid of images from the specific ward (e.g. top left = 00).

Other Scripts

Map Visualization: Javascript app using the Google Static Maps Javascript API for visualizing images to be extracted. The scripts can be found in /javascript_app/. To run the javascript app, it is recommended to run a local Python server using:

python3 -m http.server

or something similar depending on your version of Python, and then access the app on the browser at the generated local server address.

conda activate HonsProjectEnv